You are viewing a plain text version of this content. The canonical link for it is here.
Posted to yarn-issues@hadoop.apache.org by "Wangda Tan (JIRA)" <ji...@apache.org> on 2015/07/17 20:02:05 UTC
[jira] [Commented] (YARN-3937) Introducing
REMOVE_CONTAINER_FROM_PREEMPTION event to notify Scheduler and AM when a
container is no longer to be preempted
[ https://issues.apache.org/jira/browse/YARN-3937?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14631677#comment-14631677 ]
Wangda Tan commented on YARN-3937:
----------------------------------
[~sunilg],
I agree to have a separated event add to API/scheduler. And maybe add to scheduler is more important since YARN-3769 can potentially leverage it. I don't have a solid design for YARN-3769 yet, but I think if a container is removed from "to-be-preempted" list, we shouldn't do "lazy preemption" for such containers. For API changes, I'm not sure if we need it, since a container can occur on list / off list frequently, we cannot guarantee once a container is removed from list, it won't be marked again. Personally I think we can make this is an internal event first to avoid too much noises.
> Introducing REMOVE_CONTAINER_FROM_PREEMPTION event to notify Scheduler and AM when a container is no longer to be preempted
> ---------------------------------------------------------------------------------------------------------------------------
>
> Key: YARN-3937
> URL: https://issues.apache.org/jira/browse/YARN-3937
> Project: Hadoop YARN
> Issue Type: Sub-task
> Components: capacityscheduler
> Affects Versions: 2.7.1
> Reporter: Sunil G
> Assignee: Sunil G
>
> As discussed in YARN-3784, there are scenarios like few other applications released containers or same application has revoked its resource requests. In these cases, we may not have to preempt a container which would have been marked for preemption earlier.
> Introduce a new event to remove such containers if present in the to-be-preempted list of scheduler or inform AM about such a scenario.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)